What size slide show do I make

[Barry Beckham]

If you want a full frame image from a Canon 1ds Mk3 I can provide you with one, but don't forget sharpness and quality come not only from the pixels, but from the lighting, exposure, manipulation, image content, etc etc etc

[davegee]

It would be very interesting from my point of view.

So, yes please - something with plenty of clear blue sky.

DaveG

[JEB]

Hi,

For what it's worth I have looked several times on my humble HP Pavilion set at all the screen resolutions available and detect NO difference whatsoever.

John

[lathompson]

Is not an image saved at level 6 on a 22mp camera bigger than one from a 12mp camera?

I have no idea, never had reason to find out to be honest. All I know is that if you start out with great quality, that quality is still evident when the image is reduced in size. I noticed that years ago when a friend bought a really top of the range film scanner. You could see the difference in his images even at 8in by 6in email images.

With regards the Moire effect, that is a battle that each individual has to deal with and I do seem to experience it quite a bit whenever I animate an image. That is why you don't see too much animation from me. As you probably know there are lots of ways to deal with the moire effect, but they all effect image quality in some way. I have had many occasions where an animation seemed appropriate, but I had to give up the idea, because I could not get rid of the moire effect and was not prepared to accept a soft image just to use animation.

Of course the best way to defeat the moire effect is not to animate

DaveG

I have just created a crop from an original image and saved it at different Jpg compressions 12,6 etc and I cannot see any difference and I am looking at images from a 22mp camera on a brand new flat screen. I have three other people here with me and they can't tell either, so that rules out my eyesight.

Don't you think there is an obsession with this potential loss of quality? I happen to like and agree with your particular obsession, because I rather think it is the same obsession as mine. Image quality ! It is essentail in my view to what we do, but how come that in about 15 years of working with Photoshop I have not seen evidence of this loss of quality?

We save our images for a slide show, they will not be enlarged, they will not be printed, what we see is what we get. I even tried to re-create this loss of quality once by repeatingly saving the image and couldn't do it. For general advice to those who ask, why not keep the answer simple. Save a level 6 and you will retain a good balance between image quality and image size for the slide show. Many of the problems experienced by newer users are caused by the images being too large, which we have seen in this forum.

I am not suggesting that it is wrong to save images for a show at Level 12 and for those who know what they are doing and they PC will handle the files fair enough, but I do believe that it would cause more problems than it would solve if that is passed on as the best advice for all. I am sure that time will change this and in a few years we will be dropping 20meg files into our slide shows, well perhaps

I agree with you Barry. Well said.

[Barry Beckham]

Dave G

You will find an untouched (No manipulation at all from a raw file) Canon 1ds Mk3 image at http://www.beckhamdigital.co.uk/daveg.zip but, really what is the point? You are looking for evidence of a problem that most users of Photoshop never actually see in practical use. I understand that some people like the technical side of the hobby, but whatever you find it will not alter the fact that the advice we should be giving newer users of PTE is simple

Save your images as a jpg at level 6 and you won't go far wrong. What is the point in making one of the more simple tasks in the making of a slide show appear more difficult than it is. I had a slide show link sent me yesterday, but the slide show looks awful on my monitor because it is being enlarged way beyond what the author expected or even knew about. See my point about quality?

All it needed was one little box ticked and his show would have looked great. What level he chose to save his jpg images at is not irrelavent.

I hear the same arguments about unsharp mask, but my advice is always the same. Forget what you have been told about over sharpening as all of that advice is generally aimed at high resolution images destined for printing. In AV we do not print our images, we do not enlarge them and what we see at actual pixels is what we get. In addition to that, the image will be on screen for only a few seconds. So, if an image needs a little more sharpness and it is not obviously way overdone then I say give it what you think it needs.

[fh1805]

Re Jim's post (#67)...

I see very obvious pixellation in the blue sky in the Level 6 image, a little pixellation in the Level 10 and cannot discern any in the Level 12. A close study of the branches in the Level 6 and Level 12 images seems to suggest, to me, that the pixels are the same. But, in the Level 6 image, the small patches of clear blue sky between the branches show the same degree of pixellation as the large areas of clear blue sky.

My conclusion from this useful, but not necessarily scientific, test: if your image contains large areas of uniform colour, you need to check the JPEG image carefully and, if necessary, re-save at a higher compression level in order to eliminate pixellation in the uniform colour area.

Personal to Jim: I use Level 8 as my default. Any chance you could prepare and send me a Level 8 version?

regards,

Peter

[davegee]

Hi Barry,

Thanks for the file. I suppose I could have found one somewhere on the net as I did last evening with a D3x file.

I have done some checks with images from a succession of my own cameras from the last few years and the results (non-scientific; using purely my eyesight) are interesting to me. More of that later today.

DaveG

P.S. To Jim - could you tell us what camera that image came from? It tends to tie in with my test results.

[JRR]

Peter:

Quality 8 version is at :mediafire

It is 190kb.

davegee: (so we are going to start a Canon vs Nikon rivalry ?? )

Shot on a Canon 30D, colour space sRGB.

General:

The samples have been downloaded 6 times and when I read that some people could not see a difference, I went back and looked at them again. The pixellation was not as bad as it was when I made up the files, but it was still there.

When I made up the quality 8 file, I checked them all again and the pixellation on all the files were similar to what it was before. I looked at them at 100% and slightly smaller fitting my screen in photoshop.

As originally mentioned this type of picture is an exception to what is usually used, but it would not be that rare when shooting pictures outside.

I don't worry if the tree branches are pixellated as you would not notice it in the detail.

[fh1805]

Thanks, Jim. I'll take a look at level 8 in context with 6, 10 and 12.

regards,

Peter

[davegee]

I have been able over the last 24 hours to do some comparisons of the cameras I have owned over the last few years and in a totally non-scientific way try to ascertain the differences (if any) when their images are saved as JPEGs in PS.

METHODOLOGY

The images from my Nikon D70, D200 and D300 cameras were from my archives and the D3x image was downloaded from the Internet. In anticipation of JRR’s question I thought it best to stick with one brand.

Using NX2 I was able to retrospectively change the RAW data in each of the files to Neutral Picture Control and Sharpening OFF (no one sharpens their skies – do they?). No other changes were made to any of the images. All four images started off on an equal footing – the D3x image was already Neutral Picture Control and Sharpening Off.

After sending to PS each image was resized to 1920x1200 using the crop tool so that the longest side of each (landscape) image became 1920 pixels wide. After changing the MODE from 16 Bits to 8 Bits and changing the Colour Profile to sRGB I then used “Save For The Web” so that an instant comparison was available between the original image and the final JPEG at varying quality settings.

THE RESULTS

Starting with the D70 (3008x200 pixels) and concentrating on areas of continuous tone (sky at the horizon) the pixilation was severe at quality 50 (quality 6 in “Save as”?) and became acceptable at quality 80-90.

The D200 image (3872x2592 pixels) produced better results with pixilation being evident at quality 50 and being acceptable at quality 70.

The D300 image (4288x2848 pixels) was even better and produced minimal pixilation at quality 50 which disappeared at quality 60.

The D3x image (6048x4032 pixels) had no sky in it but I was able to concentrate on large areas of continuous tone and the interface between a light tone and a dark tone. I could detect absolutely no pixilation at all at quality 50.

CONCLUSIONS

It now becomes obvious to me that quality issues are not solely dependent on Photoshop’s ability to save to JPEG but the combination of different camera’s images and Photoshop’s save as JPEG function.

My “obsession” with quality began when I had the D70 and became aware of the “problem” without actually understanding the issues. I continued to produce maximum quality JPEGs when I bought my D200 and D300 as precautionary measures. Having carried out this experiment I now believe that I could relax my quality issues a little and (with my current setup) accept quality 80-100 without worrying too much.

However, there are still a lot of D70s and similar cameras out there as well as cameras with even smaller sensors. To give “blanket” advice about quality issues would be foolish without knowing what camera the person receiving the advice has. At the upper end of the quality scale saying that quality 100 (12) is advisable is fairly safe. But, with respect, saying that quality 50 (6) is acceptable without knowing the camera statistics would be, in my mind, a foolish statement.

Now, how much is that D3x again?

DaveG

[Ed Overstreet]

Thought I should throw in my two-cents' worth on Jim's test show, because my perceptions are obviously different from everyone else's.

In viewing the show in my 1600x1200 19" CRT monitor, I can't see any differences anywhere between the three levels of JPG, and in my own tests over the years I've never been convinced that JPG compression matters a darn for AV shows either on my monitor or on our Canon SX50 projector, as long as course you only save the file to JPG once and not multiple times. In fact, to keep the file size down to the website submission limit, I recently submitted for display at our club (on the SX50 at 1400x1050) a highly-detailed Level 1 JPG which looked fine to me on projection, as well as on my monitor.

Maybe it's different with LCD monitors, I don't know because I almost never use my laptop monitor.

For my eyesight anyway, this JPG-compression-level issue is Much Ado About Very Little Indeed. For the record I generally save my JPGs as Level 8 in Photoshop and at the "Good Balance" level in Nikon Capture NX2 (which is where I do maybe 90% of my editing). But the compression ratio is nothing I ever lose sleep over.

I don't see significant pixellation in the skies on Jim's test, I can almost convince myself maybe there's something there on the Level 6 image, but on repeated viewings I no longer see it. And I'm sitting with my face pretty close to the monitor screen.

No idea what it looks like to my eyes on projection, as I don't own a projector and haven't seen this test series on one.

[Ed Overstreet]

Maybe it's different with LCD monitors, I don't know because I almost never use my laptop monitor.

To be fair, I re-ran the show using my laptop's 1920x1200 17" LCD monitor. On that screen, I do detect slight evidence of pixellation on the Level 6 slide; the Level 10 and 12 slides look slighlty smoother (and equally smooth) in the blue sky areas. On re-running the show on my CRT (1600x1200 19") monitor, I really can't convince myself there is any difference among the three slides.

So I suspect part of the issue is something to do with CRT vs LCD monitors. No idea what that portends for DLP vs LCD projectors. However, the difference I see on the LCD screen is only visible because I'm looking for it. If I were to run on my LCD screen a show filled with Level 6 JPGs I very much doubt I or anyone I know would proclaim "oh my those slides are all pixellated."

The trouble with pixel-peeping is that generally differences are only visible side-by-side, and how many AV shows do any of us produce with side-by-side comparisons of variations on the same image, or sequential views of variations as in Jim's show?

I remain of the view -- Much Ado About Nothing (with a nod to Xaver)