PDA

View Full Version : BreezeBrowser vs YarcPlus


Roger_Cavanagh
20th of August 2002 (Tue), 17:58
Has anyone done a systematic comparison of these two converters? I mean wrt image quality rather than UI and other stuff.

I decided to cough up the reg fee. There are a few people, like Taz, who hang around here who have mentioned they use YP, and I came across a head-to-head review that rated BB ahead on most things except image quality. They gave 5 stars to YP and only 3 to BB.

The USP for YP is the ARF filters. Having read the draft documentation for YP, I decided to take a close look at ARF2. I take 99% of my pictures with ISO 100/200, and this was the recommendation for best image quality.

I've been running assorted images through BB and YP, producing linear TIFF files and processing using LinearSharpen (actually LinearSharpenMenu, of which more later). I was intending to produce a formal review article for rogercavanagh.com, but found it difficult to set up comparison images that would show real differences when posted to the web. So I'm posting these subjective and somewhat tentative conclusions to see what you all out there have found.

Colour: there is a definite difference in colour. There's more red in the BB image and a bit more vibrancy/brightness.

Noise: there is much less noise/artefacts in the YP images. This is especially noticeable in shadows and blurred, out of focus backgrounds.

Sharpening: we're all used to high levels of USM causing the noise to be increased to unacceptable levels. It seemed to me that the YP images (presumably because of the ARF effect on artefacts) could accept higher levels of sharpening than the BB images without affecting the noise so much. This was easier to see, if images were upsized.

Related to this, as I said, I was using LSM to process the images. Regular users of LS342 will know that it has a high quality option, which includes an upsize-downsize step to reduce jaggies. One of the refinements I added with LSM was a third quality option. This does the upsize in increments (stair interpolation). Using this Q3 option adds signficantly to image processing time, and its benefit is not apparent on every image, but sometimes the improved image detail is noticeable. I processed a number of images using this option. It seems to me that such images - both BB and YP conversions - can also accept a higher degree of sharpening before noise becomes unacceptable.

Obviously, these judgements are subjective and I'm only working with screen images at the moment, but I'd be interested to read the views of anyone else who has compared BB and YP.

Regards,

photoArne
22nd of August 2002 (Thu), 16:52
Roger,
I have only made comparison between non-linear conversions (so far) with Raw files from the D-60 and the differences are small, but not insignificant. In standard mode YP and BB are quite similar with YP having a slightly crisper rendition of the finest details. I have to go to 300% on screen to verify this.
With ARF (2) the "crispness" is certainly increased, in fact it seems to me that this is its main function. One would expect a mode named Artifact Removal Filter to remove for instance ringing or jaggies, but it does nothing of the sort, in fact it can introduce strange looking interference patterns on some subjects. On a shot of some lovely medieval houses the rows of small tiles on the roofs were at several different angels to the horisontal creating quite difficult patterns to reproduce. It looked reasonably natural in standard mode (YP and BB) but with the ARF on some parts of the roof looked "processed" and unnatural. So beware of ARF, at least for critical work I would suggest to convert in both modes and compare.
As for color, it seems to me the reds of YP are more lively going slightly in the yellow/orange direction while BB's more subdued red tends towards brownish. These differences are visible in a side by side comparison, but I'm not shure they would be noticed otherwise.
Which one to have? If ARF seems important the choice is obvious, if not other features might be decisive.
Regards
Arne

Roger_Cavanagh
22nd of August 2002 (Thu), 21:22
Arne,

Thanks for the comments. Your point about difficult patterns is very interesting. You have supported my feeling that it is not a clear cut question of ARF giving better image quality for every image. More experience I think is needed to decide which is the better default option - BB or YP with or without ARF - for batches.

WRT jaggies, I do think that Pekka's method of upsize-downsize does really help a lot, at least, on the D30 - can't speak for the D60.

One thing's for sure, I'm still using BB for sorting, reviewing, etc. The UI is much better than YP.

Regards,

pigasus
23rd of August 2002 (Fri), 03:35
Roger,

I did the initial testing on Yarc using ARF on linear conversions for Michael T. Using ARF1 followed by LS, I found that conversions using ARF1 required slightly higher sharpening in LS than without ARF1. In fact it appeared to me to be about 1/2 more required.

Michael told me that ARF2 differed from ARF1 in that it actually introduced a sharpening element. Again, doing comparisons in LS I found I thus needed less sharpening in LS with ARF2 than with no ARF.

I am now converting using Yarc and ARF1. I have avoided ARF2 as I'd rather not be scattering my sharpening around in a multitude of places.

The best combination I have found is using Yarc ARF1 with LS for 400ISO HQ, regardless of the actual ISO of the image. I like the photographic naturalness of the result. For the web, the only additional sharpening I generally apply is the slight sharpening in Pekka's Gallerysave action. For printing I apply additional sharpening with NIK.

Your idea of step interpolating the upsizing in LS sounds good. Using the 400ISO option, however, I am already operating in the realm of VERY slow. I wonder how step interpolation would affect it. :)

Regards,

Roger_Cavanagh
23rd of August 2002 (Fri), 05:00
pigasus wrote:
Your idea of step interpolating the upsizing in LS sounds good. Using the 400ISO option, however, I am already operating in the realm of VERY slow. I wonder how step interpolation would affect it. :)

Oh, no need to wonder, Sally, it's suuloooooow.... :D but it does seem to produce crisper detail in some shots.

Your point about not scattering the sharpening around is a good one.

Regards,

rkoski
27th of August 2002 (Tue), 20:37
Sorry to be a bit off topic, but is there any consensus on the Bibble Labs product compared to these two?

My understanding is that this product is just now supporting the D30/D60.

thanks,
Bob
San Jose, CA

Roger_Cavanagh
28th of August 2002 (Wed), 03:15
Bob,

Bibble appears to be more popular among Nikon users. The web site currently claims Canon support for only the D30. I did try a trial version, some time ago, but remember being wildly unimpressed with the interface.

Intermittent posts of various Canon forums about poor colour have made me disinclined to give it a second chance. And it ain't no bargain at USD99 compared to USD35 for BreezeBrowser and USD49 for YarcPlus. Still there are PC and Mac versions.

Regards,