Preparing figures for publication --PPI vs DPI
Locked
1234
1234
|
I don't accept that. Read my chapter in the Pawley book. When we
resample, we are trying to get the closest result we can to sampling the original specimen at the higher resolution. Bicubic interpolation does that, and is what we should use. We always have to remember that our digital image is not a precise representation of the microscope image, it is a set of sample points. Our task is to re-map that set of sample points to the best possible representation of the image. There are in principle, I think, even better algorithms for mapping but they are not likely to be available in your typical image processing software. Guy Optical Imaging Techniques in Cell Biology by Guy Cox CRC Press / Taylor & Francis http://www.guycox.com/optical.htm ______________________________________________ Associate Professor Guy Cox, MA, DPhil(Oxon) Australian Centre for Microscopy & Microanalysis, Madsen Building F09, University of Sydney, NSW 2006 Phone +61 2 9351 3176 Fax +61 2 9351 7682 Mobile 0413 281 861 ______________________________________________ http://www.guycox.net -----Original Message----- From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Jeremy Adler Sent: Thursday, 1 July 2010 7:33 PM To: [hidden email] Subject: Re: Preparing figures for publication --PPI vs DPI It would be simpler if journals would specify the exact width(s) of images that they can accommodate in pixels - you can treat dpi as a number of pixels. Making small adjustments to the size of images creates a difference between the image you submit and the image as published since the need to increase or decrease the number of pixels involves fabricating new pixels from the old. Ideally there should be an integer relationship between the size of the original images in pixels as it comes off the microscope and the number of pixels in the submitted image. When magnifying an image by increasing the number of pixels you should only magnify by an integer and specify (Photoshop) that interpolation is by using the nearest neighbour. Then one of your original pixels will still appear to be a single pixel when printed. If necessary pad the final image out with a border but don't give the journal an excuse for fiddling with the number of pixels. Jeremy Adler Uppsala U Sweden No virus found in this incoming message. Checked by AVG - www.avg.com Version: 9.0.830 / Virus Database: 271.1.1/2914 - Release Date: 07/01/10 04:38:00 |
 
|
JOEL B. SHEFFIELD |
Re: Preparing figures for publication --PPI vs DPI
|
|
Folks,
It seems to me that we are talking about purity here. After forty+ years in this biz, I have become quite used to the loss of quality in printed versions of my micrographs. Once, in the 60's, there was a French "Journal de Microscopie" that attempted to print high quality glossies of electron micrographs --beautiful, but unsustainable. Beyond that, it is clear that there is a significant loss of detail in printing on paper, regardless of the demands of publishers. --and forget about dynamic range! The solution that many of us have used is to show a low magnification image for orientation, with a cutout at higher magnification to illustrate the fine details. --and forget about dynamic range! What I find encouraging is the practice among some journals of posting online high resolution versions of the low-res images that they are forced to use in print. This allows one to really examine the images in detail. Unfortunately, the pdf versions of the papers generally use the print versions of the images, in order to manage file sizes, etc. As a result, I often find myself downloading not only the pdf, but also the detailed figures, for papers where this matters. If we want to approach "purity", I would join the chorus of those who ask publishers to maintain online access to the original high resolution images that we provide. In the broader scheme, though, aren't most of our published images supposed to be representative of a set of experimental results? It is only very rarely that a single picture representing a single instance of a phenomenon is worthy of publication. (I recall some virologists saying "One particle makes an article", but that's a different story>) In that case, what we are really after is to present clarity, so that the reader is able to follow our argument. Since we are already selecting images for presentation, that is, by itself, a transform of the original results, and we should be aware of that. rant over. Joel On Thu, Jul 1, 2010 at 8:31 AM, Guy Cox <[hidden email]> wrote: I don't accept that. Read my chapter in the Pawley book. When we -- Joel B. Sheffield, Ph.D Department of Biology Temple University Philadelphia, PA 19122 Voice: 215 204 8839 e-mail: [hidden email] URL: http://astro.temple.edu/~jbs |
|
|
Yes, well, it’s more than 40 years since my first
published paper too! Journal de Microscopie used photogravure – a screenless
printing process often used in art books. I have published in it.
But there is still a limit to the resolution it can present – essentially
it uses a fairly coarse-grained photographic medium as a random
screen. The original Journal of Ultrastructure Research used a screen,
but at the fabulous resolution of 200 dpi. (Several printers I’ve
talked to didn’t believe that was possible). I’ve published
in that, too. Both were well and truly above the norm. Journal
of Ultrastructural research is now, after a few incarnations, Journal of
Structural Biology but although it still does a good job with images I don’t
think it’s still using that ‘impossible’ screen.
Guy Optical Imaging Techniques in Cell Biology by Guy Cox CRC Press / Taylor & Francis http://www.guycox.com/optical.htm ______________________________________________ Associate Professor Guy Cox, MA, DPhil(Oxon) Australian Centre for Microscopy & Microanalysis, Madsen Building F09, University of Sydney, NSW 2006 Phone +61 2 9351 3176 Fax +61 2 9351
7682
Mobile 0413 281 861 ______________________________________________ From: Confocal Microscopy List
[mailto:[hidden email]] On Behalf Of JOEL B. SHEFFIELD Folks, On Thu, Jul 1, 2010 at 8:31 AM, Guy Cox <[hidden email]> wrote: I don't accept that. Read my chapter in the Pawley
book. When we
-----Original Message----- On Behalf Of Jeremy Adler It would be simpler if journals
would specify the exact width(s) of No virus found in this incoming message. Version: 9.0.830 / Virus Database: 271.1.1/2914 - Release
Date: 07/01/10
No virus
found in this incoming message. |
|
|
Jason Swedlow |
Re: Preparing figures for publication --PPI vs DPI
|
|
****COMPLETELY CONFLICTED RESPONSE******
Dear All- Rants are fine, and necessary sometimes. But it's worth considering real work and activity in the directions you are discussing. We've posted previously to these lists, on just these issues. The Journal of Cell Biology, in collaboration with Glencoe Software, Inc. and the Open Microscopy Environment, has started to build exactly the kind of applications you are discussing. Some links: The JCB DataViewer: http://jcb-dataviewer.rupress.org/ The original editorial: http://jcb.rupress.org/content/183/6/969.full A recent commentary, that includes some info on where the DataViewer is going, and some thoughts on standardized file formats: http://jcb.rupress.org/content/189/5/777.full The JCB DataViewer is built on Bio-Formats and OMERO, open source resources from the Open Microscopy Environment well known to this list. For more info: http://www.openmicroscopy.org/ Cheers, Jason DISCLAIMER: I co-founded OME and Glencoe Software. -- ************************** Wellcome Trust Centre for Gene Regulation & Expression College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dow Street Dundee DD1 5EH United Kingdom phone (01382) 385819 Intl phone: 44 1382 385819 FAX (01382) 388072 email: [hidden email] Lab Page: http://gre.lifesci.dundee.ac.uk/staff/jason_swedlow.html Open Microscopy Environment: http://openmicroscopy.org ************************** On Thu, Jul 1, 2010 at 3:16 PM, Guy Cox <[hidden email]> wrote:
|
|
|
Daniel James White |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Jay Vyas
Hi Johannes,
Here I think we are mixing up 2 separate problems. On Jul 2, 2010, at 7:01 AM, CONFOCALMICROSCOPY automatic digest system wrote: > Date: Thu, 1 Jul 2010 09:54:12 +0200 > From: "Johannes-P. Koch" <[hidden email]> > Subject: Re: Preparing figures for publication --PPI vs DPI > > Just to get it right: > > Say, I have a CLSM image with 1024x1024px, which was acquired at exc/em > 488/520nm using a 63x NA1.4 objective. The zoom factor was adjusted for > Nyquist sampling, let's say such that I have 50nm pixel size. and thats perfect;y acceptable. Good. > > I have to use these settings, since I want a whole human cell on my > image and I need the resolution as well. All good. > > If I consider now the second part of the Nyquist theorem, I would have > to use at least 2 pixels on the print for one pixel in my image - this > would be some 6.8 inches for a 300dpi printer! Here we have 2 situations; 1) Digital image in a computer = a table of numbers (pixels are point samples with a detection extent the shape of the PSF, pixel are not little squares) 2) Display of digital image on a LCD screen or in print. Here the pixels are physically little LCD elements (squares) or dots of ink on paper. So yes, to display or print that image at full resolution, one must use one pritner dot, or one LCD element to display one point sample (pixel) of the image. So the image "size" will be the number of pixels in x or y divided by the resolution( DPI) of the display or printer. Indeed, for a typical high resolution image with a large field of view, the image needs to be larger than is typically acceptable for a print version of one column width or even a full page width!!! > > > This is obviously not feasible anyway! I assume, that only something > like 256x256 px are acceptable for todays image sizes in journals. This > would imply that either I am not "allowed" to use high NA objectives > (because then my field of view is rather limited if I stick to Nyquist) Here lies the mix up. Dont even bother trying to get your large field of view high resolution (large pixel number) images printed at full resolution (1 pixel per screen pixel or print dot) They will have to be down sampled to a suitable size for printing... This should not be done with dumb methods that cause aliasing artifacts. See here: http://pacific.mpi-cbg.de/wiki/index.php/Downsample images should be gaussian smoothes first to remove high frequencies, then resampled correctly. If publishers know how to do this correctly remains unclear. Often i see badly resized images in print and PDF./ The point is this: Since the original full resolution image has too many pixels to be fit on a printed page, or in a small image in a PDF, the original image must be made available as supplemental info, so the image in the paper is just a thumbnail that send you to the real image. Requirements of printing MUST NEVER dictate how you collect the images. Certainly, printing requirements must not lead you to choose some image size (number of pixels). Forget the printing DPI requirements of the instructions for authors when it comes to image acquisition, as it does not come into play at this stage. The ONLY thing that determines the number of pixels you need for a certain field of view is the spatial resolution you wish to capture in the image. This spatial resolution can be limited/determined by the NA of the objective, or by the user saying I want to get a resolution of say 3 micrometers (so pixels should be about 1 micron apart - according to Nyqvist-Shannon) > or I have to crop excessively such that the whole context in a cell is > lost (see settings above). I suppose, you all agree that neither is a > solution. correct, both are wrong. What you need to do is give a correctly downsampled image for printing a thumbnail image, AND the original image (in the original data file format fresh off the microscope software- with meta data intact!) for inclusion oas supplemental info, or submitted to some as yet non existent public image database (like genbank or PDB but for images) > Then, the question remains for me: How can I properly transmit > such an image to a journal. The answer according to all of you is > simply, that it is not possible! it is possible, if you get the arguments above. Yuo need to give them 2 images. Original and down sampled thumbnail. > > Am I right? This is disastrous. > yes its often a disaster. cheers Dan > > > Johannes Dr. Daniel James White BSc. (Hons.) PhD Senior Microscopist / Image Visualisation, Processing and Analysis Light Microscopy and Image Processing Facilities Max Planck Institute of Molecular Cell Biology and Genetics Pfotenhauerstrasse 108 01307 DRESDEN Germany +49 (0)15114966933 (German Mobile) +49 (0)351 210 2627 (Work phone at MPI-CBG) +49 (0)351 210 1078 (Fax MPI-CBG LMF) http://www.bioimagexd.net BioImageXD http://pacific.mpi-cbg.de Fiji - is just ImageJ (Batteries Included) http://www.chalkie.org.uk Dan's Homepages https://ifn.mpi-cbg.de Dresden Imaging Facility Network dan (at) chalkie.org.uk ( white (at) mpi-cbg.de ) |
|
|
Daniel James White |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Jay Vyas
Dear Jason and All,
> > Date: Thu, 1 Jul 2010 17:19:28 +0100 > From: Jason Swedlow <[hidden email]> > Subject: Re: Preparing figures for publication --PPI vs DPI > > --0016363b9c82b21672048a55d56d > Content-Type: text/plain; charset=windows-1252 > Content-Transfer-Encoding: quoted-printable > > ****COMPLETELY CONFLICTED RESPONSE****** > > Dear All- > > Rants are fine, and necessary sometimes. But it's worth considering real > work and activity in the directions you are discussing. > > We've posted previously to these lists, on just these issues. The Journal > of Cell Biology, in collaboration with Glencoe Software, Inc. and the Open > Microscopy Environment, has started to build exactly the kind of > applications you are discussing. Some links: > > The JCB DataViewer: http://jcb-dataviewer.rupress.org/ > > The original editorial: http://jcb.rupress.org/content/183/6/969.full > > A recent commentary, that includes some info on where the DataViewer is > going, and some thoughts on standardized file formats: > http://jcb.rupress.org/content/189/5/777.full > > The JCB DataViewer is built on Bio-Formats and OMERO, open source resources > from the Open Microscopy Environment well known to this list. For more > info: http://www.openmicroscopy.org/ > > Cheers, > > Jason > Cheers, > > Jason OME and Jason's group to the rescue! This is getting very close to what we need! A couple of comments: 1) This is a great effort, but its only JCB.. right? What we really need is a public database using this kind of technology, where all publishes images go. Same model as the PDB or genbank etc. EuroBioImaging project needs to make this happen.... I hope we can all support that. OME technology used there would be a great way to go I think. 2) While one can browse the images on line in the viewer, one can not take the original images and inspect the metadata, and analyse the images however you like. Or can you also download the original images from there to your own computer? This is what PDB and genbank allow, and we also need that I strongly believe. I hope we can get the infrastructure together through the EuroBioimaging Project. Great work JCB and Jason et al! cheers Dan Dr. Daniel James White BSc. (Hons.) PhD Senior Microscopist / Image Visualisation, Processing and Analysis Light Microscopy and Image Processing Facilities Max Planck Institute of Molecular Cell Biology and Genetics Pfotenhauerstrasse 108 01307 DRESDEN Germany +49 (0)15114966933 (German Mobile) +49 (0)351 210 2627 (Work phone at MPI-CBG) +49 (0)351 210 1078 (Fax MPI-CBG LMF) http://www.bioimagexd.net BioImageXD http://pacific.mpi-cbg.de Fiji - is just ImageJ (Batteries Included) http://www.chalkie.org.uk Dan's Homepages https://ifn.mpi-cbg.de Dresden Imaging Facility Network dan (at) chalkie.org.uk ( white (at) mpi-cbg.de ) |
|
|
Knecht, David |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Guy Cox-2
Getting back to the real world, I am personally not interested in the printed page, but the image on screen from online publishing- and that invariably means pdf. With that in mind, I have a few questions:
1. Guy states that Illustrator and CorelDraw lose control of images. I don't use either, but normally use Canvas for combining graphs, text and images to make a figure. Canvas specifically does retain control of resolution of images and has bitmap image processing tools built in. Do the others not? Since Canvas no longer supports the Mac, I am looking for a new program to use and Illustrator has been suggested as an alternative. If it doesn't have this control, then it is off my list. 2. If I increase the resolution of a 512x512 image to make a montage of 1"x1" images (now 512 dpi) for the figure, I have lost no data and have not resampled, so all the data is there exactly as the original. It just looks different on the page. However, what happens when I now output that figure for the journal? They usually want 300dpi or greater tiff, eps or pdf depending on the journal. I guess the image is now being downsampled or upsampled depending, but I worry that hte mix of vector and bitmap data may be more complex and I never thought about the problem prior to this excellent discussion. 3. On the topic of miniature figures in pdf reprints, I have been periodically surprised by my ability to zoom the pdf and have lots of underlying data preserved in the tiny images of a paper. The endpoint in all cases these days will be pdf for the forseeable future. Does it matter how we output to the publisher (tiff, eps, pdf) and what they then do in terms of how much data is preserved in the final pdf. Should we be insisting on supplying a particular submission format to make this work as well as possible. Dave On Jun 30, 2010, at 8:51 AM, Guy Cox wrote: > There are a whole lot of points here. In the days of silver bromide prints there was really no possibility of access to any meaningful ‘original data’. Even if you got hold of the negatives there was no way of telling the exposure and development conditions, and hence the final gamma. The digital world gives us much more control, but also much more traceablility. > > Sadly (and I speak as a journal editor) publishers (with absolutely NO exceptions) do not understand the world of scientific digital imaging. Images in a journal are typically printed with a 133 dpi screen (150 or better if it is a high-class image-focussed journal). Now Nyquist applies here as everywhere and that means that they should give 2.3 dots to every pixel of your original image. So a 512x512 confocal image should appear ~9 inches wide. This will not happen. They want 300 dpi to give them a margin of error. Therefore the only way to get a satisfactory reproduction of your digital image is to resample to that resolution. The algorithm you use for resampling is important (see my chapter in the legendary Pawley Handbook of Confocal Microscopy) but most high-end image software will do the right thing. There is absolutely nothing wrong in doing this – it is no different to choosing the magnification on an enlarger – but in an ideal world it should not be necessary. > > However, I really do advise against importing your images into vector graphics programs such as Adobe Illustrator or Corel Draw unless they are going to be part of a really complex final graphic. Once your image is in that sort of software you have lost control of it. Bitmap software such as Photoshop or Paintshop Pro is quite capable of doing routine annotations – in vector or bitmap form – without modifying your base image. > > Guy > > Optical Imaging Techniques in Cell Biology > by Guy Cox CRC Press / Taylor & Francis > http://www.guycox.com/optical.htm > ______________________________________________ > Associate Professor Guy Cox, MA, DPhil(Oxon) > Australian Centre for Microscopy & Microanalysis, > Madsen Building F09, University of Sydney, NSW 2006 > > Phone +61 2 9351 3176 Fax +61 2 9351 7682 > Mobile 0413 281 861 > ______________________________________________ > http://www.guycox.net > > > From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Eric Scarfone > Sent: Wednesday, 30 June 2010 9:24 PM > To: [hidden email] > Subject: Re: Preparing figures for publication --PPI vs DPI > > Hej List! > I fully endorse Daniel's points in this thread. The imaging community must work towards an "ImageBank" where the original data from all published scientific image would be accessible. > > > It happens more than once that published images are so degraded that it becomes impossible to assess the author’s claims. Maybe the referees have had access to the data supporting the claims but the readers often do not. We are not anymore at the analog time when only referees could aces the original prints made by authors (which by the way was not that perfect-a-time since the original data, -the negatives- were not at all accessible!). Indeed the advent of the digital age makes it possible to look with the same ccd eye into someone else’s microscope. > > This is important for the respect of the reproducibility requirement. > This is also fundamental for the tracing of image modifications made by authors. Here this thread joins another discussion that was vivid sometimes on the list about what is allowed and not to improve images. > > A problem remains though; it is that original data really must be “original” (meaning “un-tampered with”!) > > Cheers > Eric > > > > Eric Scarfone, PhD, CNRS, > Center for Hearing and communication Research > Department of Clinical Neuroscience > Karolinska Institutet > > Postal Address: > CFH, M1:02 > Karolinska Hospital, > SE-171 76 Stockholm, Sweden > > Work: +46 (0)8-517 79343, > Cell: +46 (0)70 888 2352 > Fax: +46 (0)8-301876 > > email: [hidden email] > http://www.ki.se/cfh/ > > > ----- Original Message ----- > From: Daniel James White <[hidden email]> > Date: Wednesday, June 30, 2010 10:34 am > Subject: Re: Preparing figures for publication --PPI vs DPI > To: [hidden email] > > > Dear All, > > > > On Jun 30, 2010, at 7:02 AM, CONFOCALMICROSCOPY automatic digest > > system wrote: > > > > > > > > Date: Tue, 29 Jun 2010 10:46:53 -0400 > > > From: "JOEL B. SHEFFIELD" <[hidden email]> > > > Subject: Re: Preparing figures for publication --PPI vs DPI > > > > > > > > > The problems with letting the publisher do the work is that we > > often have a > > > particular layout in mind, and we submit complete "plates" > > rather than > > > individual images. This is one way to avoid the kinds of > > printer errors > > > that we are all familiar with. In order to comply with the > > printer's> constraints, and get the results that we want, it is > > probably safest to > > > create our own 300 dpi image that falls within the plate size of the > > > publication. > > > > In principle I agree Joel, > > but the problem really lies not really in the print versions, > > which are always going to be poor representation of a digital image... > > > > rather in the PDF version, which contain a mashed, lossy > > compressed, > > resampled and generally destroyed version of the image data. > > > > Let us not forget, out microscpes are fanct spectrometers that > > collect numerical data. > > A digital image is not analogue artwork, its just a table of > > numbers!!! > > we would never allow a text table of numbers to have its > > information destroyed like this > > on the way to it being read by a reader/reviewer, so why do we > > accept it as normal for images? > > > > We meed to change the way publication works, and make sure editors > > understand that images > > are tables of numbers and should be treated as such. > > The original non destroyed image data should always be made > > available over the web. > > > > Chemists and physicists thing cell biology etc are very very soft > > fluffy disciplines, > > and trust very little they read of out work, > > exactly for reasons like this. > > > > > > > > As one more thought, increasing numbers of journals actually do > > a very poor > > > job of producing images in print, with the idea that the printed > > image is a > > > marker for a high resolution version that is online. I would > > just ask that > > > the publishers make the the original images available, and not > > just the pdf > > > conversions of the print versions. > > > > Hear Hear!!! > > > > > > > On Behalf Of Carl Boswell > > > Sent: Tuesday, June 29, 2010 11:19 AM > > > To: [hidden email] > > > Subject: Re: Preparing figures for publication --PPI vs DPI > > > > > > =20 > > > > > > My view is that the fewer options the publisher (in reality, the > > > printer) have to modify your image, the better. Given that, I would > > > think that making all the necessary adjustments, using the tips > > already> mentioned, would give you the best chance of having your > > work appear as > > > planned in your publication. If it is left up to someone who knows > > > printing but not cell biology, you could be in for a very special > > > surprise. =20 > > > > sure, it sounds like a good plan, but in the end the image is > > always destroyed at the final stage - the printer, > > which imposes its very limited capabilities on any image. > > We can try to make sureimages look nice, but this really is only art. > > > > What a reader really needs is access to the full resolution > > original image to open in ImageJ/Fiji etc. > > > > > > > > =20 > > > > > > At the least, insist on a galley proof with the actual pictures > > that are > > > going into the paper. That way you can at least be prepared for > > what> will show up in the journal, and maybe you can have them fix > > the more > > > glaring errors. Problems with hue and contrast may be > > irritating but > > > not terminal. However, if your images have been repeatedly > > resaved as > > > jpg's, or resampled incorrectly, and the mitochondria look like > > they are > > > made from Lego's, then something needs to be said. > > > > I agree ver strongly. We must to accept badly destroyed images , > > even in print. > > Lobby your editor until you get what is right. > > > > > > > > > > > > Date: Tue, 29 Jun 2010 14:45:58 -0700 > > > From: Doug Cromey <[hidden email]> > > > Subject: Re: Preparing figures for publication --PPI vs DPI > > > > > > Carl and others, > > > > > > Don't count on the journals to know what to do. The typesetting > > from the > > > paper I cited in my earlier post was outsourced to a far away > > country, the > > > publishing was being done in an EU country, and the Editor was > > in the USA. > > > The first two galleys I saw had major JPEG artifacts in my carefully > > > prepared figures (PDF includes JPEG for the figures). > > Fortunately the > > > Editor pushed for the publisher to be less aggressive with my > > paper (5 large > > > figures) and instead of final PDF of less than 1MB, the paper > > came out at > > > 4.7MB with no appreciable artifacts. > > > > but the images are still not really good enough for someone to > > take and re do your image analysis, > > or make their own new measurements from . > > > > this is a basic requirement of scientific publication, > > and is harly ever me in our discipline, > > and this is very wrong. > > Physicists laugh at us... and biochemists. > > > > there is no technical reason that raw data can not be made > > available on line > > even big screens, > > its just missing infrastructure that needs to be put in place. > > > > Where is our biological image equivalent of the PDB database, the > > GENBANK database, > > all the other databases other disciplines have. > > Where is ours? We need to get that funded and organised through eg > > Euro-Bioimaging project. > > http://www.eurobioimaging.eu/ > > and an USA / asia pacific equivlaent > > > > > > > > > > > > The "art" department at most publishers seems to mostly consist > > of people > > > who are accustomed to graphic design, not science. I'm afraid > > that Daniel's > > > complaints about being forced to do the publisher's work are > > unrealistic,> unfortunately we often need to be smarter than the > > publisher, even if that > > > should not have to be our job. > > > > Or we need to force our publishers to get smart about the kind of > > images we give them. > > and how to treat them. > > > > I would suggest we be proactive here, rather than defeatist. > > No matter how hard we try to make a well formatted plate for the > > publisher, > > it will most likely still be mashed by the time the pdf arrived on > > a readers screen. > > > > We need to educate the publishing system of our needs. > > I suggest using the argument that our images are 2D/3xD tables of > > spectroscopic measurement NUMBERS, > > and should be treated as such, not as photos - which they are NOT. > > > > > > cheers > > > > Dan > > > > > > > > > > > > > > Doug > > > > > > > > > > > > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > > > > > Douglas W. Cromey, M.S. - Assistant Scientific Investigator > > > > > > Dept. of Cell Biology & Anatomy, University of Arizona > > > > > > 1501 N. Campbell Ave, Tucson, AZ 85724-5044 USA > > > > > > > > > > > > office: AHSC 4212 email: [hidden email] > > > > > > voice: 520-626-2824 fax: 520-626-2097 > > > > > > > > > > > > http://swehsc.pharmacy.arizona.edu/exppath/ > > > > > > Home of: "Microscopy and Imaging Resources on the WWW" > > > > > > Dr. Daniel James White BSc. (Hons.) PhD > > Senior Microscopist / Image Visualisation, Processing and Analysis > > Light Microscopy and Image Processing Facilities > > Max Planck Institute of Molecular Cell Biology and Genetics > > Pfotenhauerstrasse 108 > > 01307 DRESDEN > > Germany > > > > +49 (0)15114966933 (German Mobile) > > +49 (0)351 210 2627 (Work phone at MPI-CBG) > > +49 (0)351 210 1078 (Fax MPI-CBG LMF) > > > > http://www.bioimagexd.net BioImageXD > > http://pacific.mpi-cbg.de Fiji - is just ImageJ > > (Batteries Included) > > http://www.chalkie.org.uk Dan's Homepages > > https://ifn.mpi-cbg.de Dresden Imaging Facility > > Networkdan (at) chalkie.org.uk > > ( white (at) mpi-cbg.de ) > > > No virus found in this incoming message. > Checked by AVG - www.avg.com > Version: 9.0.830 / Virus Database: 271.1.1/2914 - Release Date: 06/30/10 04:35:00 > Dr. David Knecht Department of Molecular and Cell Biology Co-head Flow Cytometry and Confocal Microscopy Facility U-3125 91 N. Eagleville Rd. University of Connecticut Storrs, CT 06269 860-486-2200 860-486-4331 (fax) |
|
|
Dave Tieman |
Re: Preparing figures for publication --PPI vs DPI
|
|
On 7/2/2010 1:06 PM, David Knecht wrote:
> 1. Guy states that Illustrator and CorelDraw lose control of images. I was a little puzzled at how someone would "lose control of images" in these programs. Any image that you import to the native format will keep all of its resolution until you export it to a bitmap, at which time the illustration will be resampled at the resolution specified. I assume that the loss of control is because the resampling is defined over a specific page area, rather than over a particular number of pixels: this makes it a bit trickier to import a 1024x1024 image and have the pixels correspond in some straightforward way to the pixels of a final tif, but it makes it easy enough to prepare the figure that the publisher is requesting--e.g., 3.75" at 300dpi. There are certainly situations where someone might prefer PhotoShop with a dimension of 1125px, in which case they would "lose control" of the page measurement. Page description vs. pixel peaking. > Do the others not? Since Canvas no longer supports the Mac, I am > looking for a new program to use and Illustrator has been suggested as > an alternative. If it doesn't have this control, then it is off my list. CorelDraw has bitmap processing through its integration with Corel PhotoPaint, BUT it has not been supported for the Mac since version 11, so that is a no go. You may, or may not, like Adobe's way, which will encourage you to use PhotoShop in conjuction with Illustrator. > 2. If I increase the resolution of a 512x512 image to make a montage > of 1"x1" images (now 512 dpi) for the figure, I have lost no data and > have not resampled, so all the data is there exactly as the original. > It just looks different on the page. However, what happens when I now > output that figure for the journal? They usually want 300dpi or > greater tiff, eps or pdf depending on the journal. I guess the image > is now being downsampled or upsampled depending, Correct and as it would be in CorelDraw or Illustrator. > but I worry that hte mix of vector and bitmap data may be more complex the mix is treated the same, except that the vectors are being rasterized into the image, rather than just resampled. That is more complex than bicubic smoothing. > 3. On the topic of miniature figures in pdf reprints, I have been > periodically surprised by my ability to zoom the pdf and have lots of > underlying data preserved in the tiny images of a paper. The endpoint > in all cases these days will be pdf for the forseeable future. Does > it matter how we output to the publisher (tiff, eps, pdf) and what > they then do in terms of how much data is preserved in the final pdf. > Should we be insisting on supplying a particular submission format to > make this work as well as possible. Several people have suggeested that pdf files contain low resolution jpg images. AFAIK, that is not because they are pdfs, but because of decisions made for practical (or impractical, depending on your viewpoint) reasons. For instance, in the aforementioned CorelDraw's Publish-to-PDF dialog, allowable bitmap compression types are NONE, LZW, JPEG, ZIP and JPG2, and any lossy compression can be adjusted for quality. Bitmap downsampling is optional and the dpi for that sampling can be specified. Full pdf creation software should provide such options. Whether the huge file sizes that may result are practical for readers or the publishers is another matter. dave -- David G. Tieman Technology Coordinator Department of Biological Sciences Ph: (518) 442-4317 |
|
|
Andreas Bruckbauer |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Daniel James White
Hi,
my thoughts t this very interesting discussion: 1) Applying Nyquist twice (once for taking the image and then for printing/resampling it) seems too much too me, just think about the real sample, how it can be imaged by the lens and how good it might be represented in the final printed or displayed image. Keep in mind that your computer monitor has an even worse resolution than the 300 dpi printing (mine has 110 dpi) , so the image will have to be even bigger on the screen. 2) I actually don't want to examine the whole 1024 x 1024 image for every image for every paper i read, and certainly don't want to to apply all the image processing steps myself on the original data. I want to see the interpretation of the authors and am much more happy with the low resolution overview and a suitable sized zoom in to show me the interesting structures. Instead of creating giant databases with original files, one can e-mail the author to ask for the files or repeat the measurement. 3) I am more concerned with the compression used when producing a pdf file. While the image size in dots might be right, the jpeg compression can seriously mess up small features and colors can be changed as well. I would be more in favour that journals keep processed but uncompressed high resolution files on their webpage instead of the raw data. best wishes Andreas -----Original Message-----
From: Daniel James White <[hidden email]> To: [hidden email] Sent: Fri, 2 Jul 2010 10:04 Subject: Re: Preparing figures for publication --PPI vs DPI Hi Johannes, |
|
|
George McNamara |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Jason Swedlow
Hi Jason,
I enjoyed your commentary - thanks for the link. Especially nice to see you support for open data access. Would help if the JCB abstract and supplemental data pages linked to the article's DataViewer page. Your lab's Jaqaman et al articles does neither, http://jcb.rupress.org/content/188/5/665.abstract and http://jcb.rupress.org/content/188/5/665/suppl/DC1 (the DataViewer page does link to the article). Your commentary figure 2 objective settings list would benefit from objective lens type (plan apo, etc), magnification, and especially numerical aperture. Would also help to have condenser settings. Within channels, I have never seen any benefit to publishing confocal microscope pinhole size in um - how about specifying Airy units (and/or making sure the metadata has all the parameters needed to convert between the two). As for this thread ppi is moot: the solution is to publish all the data online. Sincerely, George At 12:19 PM 7/1/2010, you wrote: ****COMPLETELY CONFLICTED RESPONSE****** George McNamara, Ph.D. Image Core Manager Analytical Imaging Core Facility University of Miami, Miller School of Medicine Miami, FL 33136 [hidden email] [hidden email] 305-243-8436 office http://www.sylvester.org/AICF (Analytical Imaging Core Facility) http://www.sylvester.org/AICF/pubspectra.zip (the entire 2000+ spectra .xlsx file is in the zip file) http://home.earthlink.net/~geomcnamara |
|
|
George McNamara |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Daniel James White
Hi Dan,
Read the article - page 2 says the next update will enable users to download data. Dear listservites (or at least those interested in this thread), Most journal's don't care how long the supplemental pdf is - see http://diabetes.diabetesjournals.org/content/59/4/947/suppl/DC1 for a recent paper whose supplemental file has 37 figures - about 32 of them images. Sincerely, George At 05:26 AM 7/2/2010, you wrote: >Dear Jason and All, > > > > > > Date: Thu, 1 Jul 2010 17:19:28 +0100 > > From: Jason Swedlow <[hidden email]> > > Subject: Re: Preparing figures for publication --PPI vs DPI > > > > --0016363b9c82b21672048a55d56d > > Content-Type: text/plain; charset=windows-1252 > > Content-Transfer-Encoding: quoted-printable > > > > ****COMPLETELY CONFLICTED RESPONSE****** > > > > Dear All- > > > > Rants are fine, and necessary sometimes. But it's worth considering real > > work and activity in the directions you are discussing. > > > > We've posted previously to these lists, on just these issues. The Journal > > of Cell Biology, in collaboration with Glencoe Software, Inc. and the Open > > Microscopy Environment, has started to build exactly the kind of > > applications you are discussing. Some links: > > > > The JCB DataViewer: http://jcb-dataviewer.rupress.org/ > > > > The original editorial: http://jcb.rupress.org/content/183/6/969.full > > > > A recent commentary, that includes some info on where the DataViewer is > > going, and some thoughts on standardized file formats: > > http://jcb.rupress.org/content/189/5/777.full > > > > The JCB DataViewer is built on Bio-Formats and OMERO, open source resources > > from the Open Microscopy Environment well known to this list. For more > > info: http://www.openmicroscopy.org/ > > > > Cheers, > > > > Jason > > Cheers, > > > > Jason > >OME and Jason's group to the rescue! >This is getting very close to what we need! > >A couple of comments: > >1) This is a great effort, but its only JCB.. right? >What we really need is a public database using this kind of technology, >where all publishes images go. Same model as the PDB or genbank etc. >EuroBioImaging project needs to make this happen.... I hope we can >all support that. >OME technology used there would be a great way to go I think. > >2) While one can browse the images on line in the viewer, >one can not take the original images and inspect the metadata, >and analyse the images however you like. >Or can you also download the original images from there to your own computer? >This is what PDB and genbank allow, and we also need that I strongly believe. >I hope we can get the infrastructure together through the >EuroBioimaging Project. > >Great work JCB and Jason et al! > >cheers > >Dan > > > > >Dr. Daniel James White BSc. (Hons.) PhD >Senior Microscopist / Image Visualisation, Processing and Analysis >Light Microscopy and Image Processing Facilities >Max Planck Institute of Molecular Cell Biology and Genetics >Pfotenhauerstrasse 108 >01307 DRESDEN >Germany > >+49 (0)15114966933 (German Mobile) >+49 (0)351 210 2627 (Work phone at MPI-CBG) >+49 (0)351 210 1078 (Fax MPI-CBG LMF) > >http://www.bioimagexd.net BioImageXD >http://pacific.mpi-cbg.de Fiji - is just ImageJ >(Batteries Included) >http://www.chalkie.org.uk Dan's Homepages >https://ifn.mpi-cbg.de Dresden Imaging Facility Network >dan (at) chalkie.org.uk >( white (at) mpi-cbg.de ) George McNamara, Ph.D. Image Core Manager Analytical Imaging Core Facility University of Miami, Miller School of Medicine Miami, FL 33136 [hidden email] [hidden email] 305-243-8436 office http://www.sylvester.org/AICF (Analytical Imaging Core Facility) http://www.sylvester.org/AICF/pubspectra.zip (the entire 2000+ spectra .xlsx file is in the zip file) http://home.earthlink.net/~geomcnamara |
|
|
Andreas Bruckbauer |
Re: Preparing figures for publication --PPI vs DPI
|
|
Hi George,
>Most journal's don't care how long the supplemental pdf is cell press has tightened its guidlines: The total number of supplemental data items of all types (figures, tables, movies and other) per paper may not exceed two times the number of figures and tables in the main paper. For example, a paper with 7 main figures can have up to 14 supplemental items total, of which up to 7 may be figures. As one figure usualy contains several images you could not submit your original files as supplmentary data to cell press jornals (except biophys j. maybe). Furthermore think about the disk space and download times for multi color 4D data sets of 1GB or more each. I think keeping the original files should be responsibility of the author and not the journal. It will only increase publishing prices further. best wishes Andreas Sincerely, George At 05:26 AM 7/2/2010, you wrote: >Dear Jason and All, > > > > > > Date: Thu, 1 Jul 2010 17:19:28 +0100 > > From: Jason Swedlow <[hidden email]> > > Subject: Re: Preparing figures for publication --PPI vs DPI > > > > --0016363b9c82b21672048a55d56d > > Content-Type: text/plain; charset=windows-1252 > > Content-Transfer-Encoding: quoted-printable > > > > ****COMPLETELY CONFLICTED RESPONSE****** > > > > Dear All- > > > > Rants are fine, and necessary sometimes. But it's worth considering real > > work and activity in the directions you are discussing. > > > > We've posted previously to these lists, on just these issues. The Journal > > of Cell Biology, in collaboration with Glencoe Software, Inc. and the Open > > Microscopy Environment, has started to build exactly the kind of > > applications you are discussing. Some links: > > > > The JCB DataViewer: http://jcb-dataviewer.rupress.org/ > > > > The original editorial: http://jcb.rupress.org/content/183/6/969.full > > > > A recent commentary, that includes some info on where the DataViewer is > > going, and some thoughts on standardized file formats: > > http://jcb.rupress.org/content/189/5/777.full > > > > The JCB DataViewer is built on Bio-Formats and OMERO, open source resources > > from the Open Microscopy Environment well known to this list. For more > > info: http://www.openmicroscopy.org/ > > > > Cheers, > > > > Jason > > Cheers, > > > > Jason > >OME and Jason's group to the rescue! >This is getting very close to what we need! > >A couple of comments: > >1) This is a great effort, but its only JCB.. right? >What we really need is a public database using this kind of technology, >where all publishes images go. Same model as the PDB or genbank etc. >EuroBioImaging project needs to make this happen.... I hope we can >all support that. >OME technology used there would be a great way to go I think. > >2) While one can browse the images on line in the viewer, >one can not take the original images and inspect the metadata, >and analyse the images however you like. >Or can you also download the original images from there to your own computer? >This is what PDB and genbank allow, and we also need that I strongly believe. >I hope we can get the infrastructure together through the >EuroBioimaging Project. > >Great work JCB and Jason et al! > >cheers > >Dan > > > > >Dr. Daniel James White BSc. (Hons.) PhD >Senior Microscopist / Image Visualisation, Processing and Analysis >Light Microscopy and Image Processing Facilities >Max Planck Institute of Molecular Cell Biology and Genetics >Pfotenhauerstrasse 108 >01307 DRESDEN >Germany > >+49 (0)15114966933 (German Mobile) >+49 (0)351 210 2627 (Work phone at MPI-CBG) >+49 (0)351 210 1078 (Fax MPI-CBG LMF) > >http://www.bioimagexd.net BioImageXD >http://pacific.mpi-cbg.de Fiji - is just ImageJ >(Batteries Included) >http://www.chalkie.org.uk Dan's Homepages >https://ifn.mpi-cbg.de Dresden Imaging Facility Network >dan (at) chalkie.org.uk >( white (at) mpi-cbg.de ) George McNamara, Ph.D. Image Core Manager Analytical Imaging Core Facility University of Miami, Miller School of Medicine Miami, FL 33136 [hidden email] [hidden email] 305-243-8436 office http://www.sylvester.org/AICF (Analytical Imaging Core Facility) http://www.sylvester.org/AICF/pubspectra.zip (the entire 2000+ spectra .xlsx file is in the zip file) http://home.earthlink.net/~geomcnamara |
|
|
Daniel James White |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Jay Vyas
Hi Dave,
On Jul 3, 2010, at 7:01 AM, CONFOCALMICROSCOPY automatic digest system wrote: > Date: Fri, 2 Jul 2010 13:06:04 -0400 > From: David Knecht <[hidden email]> > Subject: Re: Preparing figures for publication --PPI vs DPI > > Getting back to the real world, I am personally not interested in the = > printed page, I think thats the right attitude. These printed images are no more than thumbnails. > but the image on screen from online publishing- and that = > invariably means pdf. With that in mind, I have a few questions: > 1. Guy states that Illustrator and CorelDraw lose control of images. I = > don't use either, but normally use Canvas for combining graphs, text and = > images to make a figure. Canvas specifically does retain control of = > resolution of images and has bitmap image processing tools built in. Do = > the others not? Since Canvas no longer supports the Mac, I am looking = > for a new program to use and Illustrator has been suggested as an = > alternative. If it doesn't have this control, then it is off my list. I think here we are missing the point a little. The images have their own intrinsic scaling, and thats hopefully in the meta data, and displayed in the image as a scale bar. Surely the images as supplied to the publishers dont need to be laid out or rescaled themselves by the author. Rather one can tell the publisher the suggested layout (which is often changed anyway as we dont always send the figuured laid out as the y will be in the final PDF anyway) I mean a "drawing" or plan of the figure, with info showing which image goes where. Its the publishers job to take the images and make that work in print/PDF > > 2. If I increase the resolution of a 512x512 image to make a montage of = > 1"x1" images (now 512 dpi) for the figure, I have lost no data and have = > not resampled, so all the data is there exactly as the original. It = > just looks different on the page. Yes, and up to that point all is good - but we have lost the meta data!!! The reader has no idea what the imaging parameters where... maybe there is a scale bar... but thats about it. The reviewer and the reader need that metadata to make sense of the image. > However, what happens when I now = > output that figure for the journal? They usually want 300dpi or greater = > tiff, eps or pdf depending on the journal. I guess the image is now = > being downsampled or upsampled depending, but I worry that hte mix of = > vector and bitmap data may be more complex and I never thought about the = > problem prior to this excellent discussion. the result is the image is rescales and information is lost. More often than not JPEG compression is also used, so the images in the final PDF as garbage and no use for re analysis. That who original data must be made available as supp. info, since the printed/PDF image can only ever be thumbnails practically speaking. > > 3. On the topic of miniature figures in pdf reprints, I have been = > periodically surprised by my ability to zoom the pdf and have lots of = > underlying data preserved in the tiny images of a paper. it might "look" ok, but if you grab the images and inspect it carefully, it might well be resampled and losst compressed. How should the csual reader know? Here is how: http://pacific.mpi-cbg.de/wiki/index.php/Detect_Information_Loss > The endpoint = > in all cases these days will be pdf for the forseeable future. Does it = > matter how we output to the publisher (tiff, eps, pdf) and what they = > then do in terms of how much data is preserved in the final pdf. Should = > we be insisting on supplying a particular submission format to make this = > work as well as possible. I think we must assume the figured in a PDF are low quality thumbnails, and we expect to lookj at the online supp info, or better at an image database entry to see the original datafiles, with meta data intact, and all image processing steps described. cheers Dan Dr. Daniel James White BSc. (Hons.) PhD Senior Microscopist / Image Visualisation, Processing and Analysis Light Microscopy and Image Processing Facilities Max Planck Institute of Molecular Cell Biology and Genetics Pfotenhauerstrasse 108 01307 DRESDEN Germany +49 (0)15114966933 (German Mobile) +49 (0)351 210 2627 (Work phone at MPI-CBG) +49 (0)351 210 1078 (Fax MPI-CBG LMF) http://www.bioimagexd.net BioImageXD http://pacific.mpi-cbg.de Fiji - is just ImageJ (Batteries Included) http://www.chalkie.org.uk Dan's Homepages https://ifn.mpi-cbg.de Dresden Imaging Facility Network dan (at) chalkie.org.uk ( white (at) mpi-cbg.de ) |
|
|
Daniel James White |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Jay Vyas
Hi Andreas,
thanks for chipping into this interesting discussion! Hi listers, I hope you aren't getting too bored of the sound of my voice.... On Jul 4, 2010, at 7:01 AM, CONFOCALMICROSCOPY automatic digest system wrote: > > Date: Sat, 3 Jul 2010 05:19:29 -0400 > From: Andreas Bruckbauer <[hidden email]> > Subject: Re: Preparing figures for publication --PPI vs DPI > > ----------MB_8CCE8A7F95F9522_1580_25870_Webmail-m109.sysops.aol.com > Content-Transfer-Encoding: quoted-printable > Content-Type: text/plain; charset="us-ascii" > > Hi, > my thoughts t this very interesting discussion: > > 1) Applying Nyquist twice (once for taking the image and then for printing= > /resampling it) seems too much too me, just think about the real sample,= > how it can be imaged by the lens and how good it might be represented in= > the final printed or displayed image. Keep in mind that your computer mon= > itor has an even worse resolution than the 300 dpi printing (mine has 110= > dpi) , so the image will have to be even bigger on the screen. Your point are good, and I agree. A big problem is that the information is digital, and print is analog and also for other reasons has a hard time making a good job of representing the image as well as a computer screen. Print is mostly dead anyway... and from a digital image point of view - its no way to "only" publish image data... which is just a list of numbers... and these numbers should be made avaialble for download so readers can analyse the data as they wish,m as they can for gene sequences, protein structures, or any other table of numbers. > > 2) I actually don't want to examine the whole 1024 x 1024 image for every= > image for every paper i read, Maybe you dont, but I do, if i am working in that particular field, and I want to see how they got to their conclusions, by repeating their analysis. This is a basic requirement of scientific publication which is sadly ignored in our field. I think we must change that - to avoid the ridicule of the "harder sciences" and force us to "do the right thing" Physicists etc. dont take out work seriously because we ignore the basic rules, and thus appear to be doing "fluffy science" - no better then holiday snaps. > and certainly don't want to to apply all th= > e image processing steps myself on the original data. maybe you dont, but I do! > I want to see the in= > terpretation of the authors and am much more happy with the low resolution= > overview and a suitable sized zoom in to show me the interesting structur= > es. Foe sure, we are interested as readers to hear their side of the story, how they interpret the data they have. BUT - we must also be allowed to take the raw data and see of we come to the same conclusion. This is a basic requirement. It is true in gene sequencing, protein structure... every thing else quantitative - except biological imaging. Thats bad for us. We must change it. It is simply not enough to rely on a few tiny thumbnail images, an then blindly believe (or not) the authors analysis and conclusions with no way of independent verification. You might not want to bother with that, but others will and do. > Instead of creating giant databases with original files, one can e-mai= > l the author to ask for the files or repeat the measurement. If you ever tried that, you probably found that is rather like asking for plasmids, clones or other material. You most often get no reply, and even when you do get some t hing in the post, its often not what you thought it was. There is no pressure on the author to be honest here, and sadly many aren't. Original image data must go in a publicly accessible database, so readers can get it and analyze it themselves if they want. Look at it this way - if i make a protein crystal structure for a systems that cures cancer... and i try to publish it in nature, but dont make the PDB/atomic coordinates and X-ray data available, I will be told politely to go boil my head. Why is this not also true. for us in biological imaging? > > 3) I am more concerned with the compression used when producing a pdf file= > . While the image size in dots might be right, the jpeg compression can se= > riously mess up small features and colors can be changed as well. Lossy compression a very bad thing. The images even look different than the originals.... never mind even tying to analyze them... > I would= > be more in favour that journals keep processed but uncompressed high reso= > lution files on their webpage instead of the raw data. Processed data might be ok, if the author provided the precise information on how the raw pixel values were manipulated to get to the result image. This is very rarely the case. The only sensible way, and the way it is done elsewhere, and should be done here is that: 1) The raw data is available, 2) The image processing steps are described fully (using NO black/magic boxes - so it repeatable by all without any special closed software) How else can you understand the image you are presented with? Without these things the image is purely art and meaningless for quantitative science. The whole idea of "I just want to see what it looks like - I don't need to measure anything" is a chronic sickness in our field - Its a reason that hard scientists don't take us seriously. The above statement is a poor excuse for simply not being bothered to actually do the experiment and measure something. An example I heard yesterday from an very intelligent and careful young researcher: I measured loads of stuff by FCS, but this thing was not measurable that way, and I just wanted to see if it was cytoplasmic or not. So I just took an image of it to see. I didn't need to measure anything for that - it looks like its in the cytoplasm." One hears this kind of argument very often, and its easy to fall into that way of thinking. But - its a big mistake. The resulting image might be pretty, and shows some stuff in a place that looks like it might be the cytoplasm, but we still know nothing! The right, and simple experiment is to take a "known" cytoplasmic marker , and look for spatial correlation between it and the stuff you are interested in. Now we measured something, and got a result. Just looking at the image, we might trick ourselves into thinking that we know something, but we do not. We must release ourselved from the straight jacket of treating digital images as "pretty photos" and get into the habit of thinking like spectroscopists instead. The image is information - a bunch if numbers - we should treat it that way - measure something - dont "just look at it". cheers Dan Dr. Daniel James White BSc. (Hons.) PhD Senior Microscopist / Image Visualisation, Processing and Analysis Light Microscopy and Image Processing Facilities Max Planck Institute of Molecular Cell Biology and Genetics Pfotenhauerstrasse 108 01307 DRESDEN Germany +49 (0)15114966933 (German Mobile) +49 (0)351 210 2627 (Work phone at MPI-CBG) +49 (0)351 210 1078 (Fax MPI-CBG LMF) http://www.bioimagexd.net BioImageXD http://pacific.mpi-cbg.de Fiji - is just ImageJ (Batteries Included) http://www.chalkie.org.uk Dan's Homepages https://ifn.mpi-cbg.de Dresden Imaging Facility Network dan (at) chalkie.org.uk ( white (at) mpi-cbg.de ) |
|
|
Martin Wessendorf-2 |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Andreas Bruckbauer
Dear Andreas--
Andreas Bruckbauer wrote: > As one figure usualy contains several images you could not submit your > original files as supplmentary data to cell press jornals (except > biophys j. maybe). Furthermore think about the disk space and download > times for multi color 4D data sets of 1GB or more each. I think keeping > the original files should be responsibility of the author and not the > journal. It will only increase publishing prices further. I see your point about publication costs. However, if supplemental data form an important part of the evidence, it presumably has undergone peer review like the remainder of the paper. At that point the reviewers or editors have the option to say, "No, this figure (or 4D dataset) really isn't necessary," and it should not be included. However, if they do see the value of the data to that publication and to their journal, then I think they have taken on the obligations of a data library. People can continue to ponder a paper journal for as long as the paper lasts. One can only read the supplemental data section for as long as the *server* lasts. Anyone know how journals are dealing with this issue? Martin Wessendorf -- Martin Wessendorf, Ph.D. office: (612) 626-0145 Assoc Prof, Dept Neuroscience lab: (612) 624-2991 University of Minnesota Preferred FAX: (612) 624-8118 6-145 Jackson Hall, 321 Church St. SE Dept Fax: (612) 626-5009 Minneapolis, MN 55455 e-mail: [hidden email] |
|
|
Daniel James White |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Jay Vyas
Hi Martin and all,
>> > > I see your point about publication costs. However, if supplemental data > form an important part of the evidence, it presumably has undergone peer > review like the remainder of the paper. At that point the reviewers or > editors have the option to say, "No, this figure (or 4D dataset) really > isn't necessary," and it should not be included. However, if they do > see the value of the data to that publication and to their journal, then > I think they have taken on the obligations of a data library. People > can continue to ponder a paper journal for as long as the paper lasts. > One can only read the supplemental data section for as long as the > *server* lasts. > > Anyone know how journals are dealing with this issue? I dont know what they are thinking right now.... and opinions probably vary a great deal... and for sure they will not want to take on the expensive burden of running data servers for the original data. I also agree that they should not have to do this. But the author should not either! What we need is out own image database, like the genbank or PDB etc we have to design it and get it funded, probably via the eurobioimaging project and other similar frameworks. Its up to us to organise it, and get the central funding set up for it, as per genbank, expasy, PDB etc. OME should provide the building blocks needed for software requirements for this. cheers Dan > > Martin Wessendorf > -- > Martin Wessendorf, Ph.D. office: (612) 626-0145 > Assoc Prof, Dept Neuroscience lab: (612) 624-2991 > University of Minnesota Preferred FAX: (612) 624-8118 > 6-145 Jackson Hall, 321 Church St. SE Dept Fax: (612) 626-5009 > Minneapolis, MN 55455 e-mail: [hidden email] Dr. Daniel James White BSc. (Hons.) PhD Senior Microscopist / Image Visualisation, Processing and Analysis Light Microscopy and Image Processing Facilities Max Planck Institute of Molecular Cell Biology and Genetics Pfotenhauerstrasse 108 01307 DRESDEN Germany +49 (0)15114966933 (German Mobile) +49 (0)351 210 2627 (Work phone at MPI-CBG) +49 (0)351 210 1078 (Fax MPI-CBG LMF) http://www.bioimagexd.net BioImageXD http://pacific.mpi-cbg.de Fiji - is just ImageJ (Batteries Included) http://www.chalkie.org.uk Dan's Homepages https://ifn.mpi-cbg.de Dresden Imaging Facility Network dan (at) chalkie.org.uk ( white (at) mpi-cbg.de ) |
|
|
Jason Swedlow |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by Daniel James White
Dear All-
Sorry for the radio silence. Was on holiday, with an email embargo. On Fri, Jul 2, 2010 at 10:26 AM, Daniel James White <[hidden email]> wrote: -- Dear Jason and All, ...snip...
Well, it's a start and a work in progress. The team at JCB, especially Mike Rossner and Emma Hill, have also made great contributions to this effort.
So far. We are of course working to get the funding to make a full resource for original image data in all life sciences journals. What we really need is a public database using this kind of technology, We'll see.
Or can you also download the original images from there to your own computer? Not yet, but we are working on this. Please understand-- alot of tech has to be built, and we are doing this, in steps, in collaboration with the JCB. We'll update the list when this is available. We've been asked repeatedly for a programmatic API for data access. Again, in the works. Our plan is to base this on the OMERO API. This is what PDB and genbank allow, and we also need that I strongly believe. Definitely, we fully agree. I hope we can get the infrastructure together through the EuroBioimaging Project. That is certainly one avenue. We are trying many ways-- we want to extend the DataViewer as soon as we can: more types of data, more journals, and more functionality. Great work JCB and Jason et al! Thanks for your support. Cheers, Jason ************************** Wellcome Trust Centre for Gene Regulation & Expression College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dow Street Dundee DD1 5EH United Kingdom phone (01382) 385819 Intl phone: 44 1382 385819 FAX (01382) 388072 email: [hidden email] Lab Page: http://gre.lifesci.dundee.ac.uk/staff/jason_swedlow.html Open Microscopy Environment: http://openmicroscopy.org ************************** |
|
|
Jason Swedlow |
Re: Preparing figures for publication --PPI vs DPI
|
|
In reply to this post by George McNamara
Hi George-
On Mon, Jul 5, 2010 at 6:57 PM, George McNamara <[hidden email]> wrote:
Interesting suggestion. As you can see, the links to the data are part of the figures. That made sense, since the original data is usually associated with the figures (but not always). I'll pass your idea along.
You are correct-- we don't support condenser very well. We'll work on that.
Yes, well, the devil is in the details. Remember-- we don't have access to the proprietary software-- just the metadata. It turns out that in some cases, the proprietary files actually store the size in µm, not in Airy units (they convert it in their software). We were hesitant to convert to Airy units, as we have to assume that all other metadata are correct. It's tricky. Thanks for the feedback. Cheers, Jason
-- ************************** Wellcome Trust Centre for Gene Regulation & Expression College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dow Street Dundee DD1 5EH United Kingdom phone (01382) 385819 Intl phone: 44 1382 385819 FAX (01382) 388072 email: [hidden email] Lab Page: http://gre.lifesci.dundee.ac.uk/staff/jason_swedlow.html Open Microscopy Environment: http://openmicroscopy.org ************************** |
|
|
In reply to this post by Jason Swedlow
At ASU we are looking into using underwriters,
such as REMI, to cover our large equipment service contracts in order to
cut down on costs. Does anyone have experience with using this type of service?
Do you have any cautions or recommendations?
Thanks!
Page
Page Baluch |
|
|
Boswell, Carl A - (cboswell) |
Re: Service Contract underwriters
|
|
Hi Page,
Their website says: "THE REMI GROUP does not repair or maintain
equipment...". What does the service provide? How is cost
reduced?
Thanks,
c
Carl A. Boswell, Ph.D.
Molecular and Cellular Biology University of Arizona 520-954-7053 FAX 520-621-3709
|
«
Return to Confocal Microscopy List
|
1 view|%1 views
| Free forum by Nabble | Edit this page |