model: Wolverine Titan 8 in 1 firmware version: FS663A.F2L44aZS.RM19.015T This is the FWFS601A.bin firmware file downloadable from Wolverine's web site as of August 1, 2018 and is contained in the file F2D_Titan_Firmware_to_reduce_yellow.zip. This slide to digital converter has many features which make it easy and productive to use. Unfortunately, the blue channel appears to have a defect that causes bright areas to take on a yellow color. While it may be possible to mask this defect by post processing the image, it is never possible to remove the defect. I ended up returning the converter because I could not accept the defect. The features which work well are: 1) It is possible to convert upwards of 200 35mm slides per hour. The 35mm slide holder is first inserted into the convert. The, you insert the slides into the holder on the left one by one, center the image, convert the image to digital by pressing two buttons in sequence, and then remove the slide which sticks out on the left. There are typically 3 or 2 slides in the holder at any one time. Inserting a slide on the right pushes all the slides to the left so the leftmost slide can be removed. 2) The image conversion process would typically take only 2 to 3 seconds. After converting a cassette of 100 slides, I would simply take the 32GB SDHC card over to my PC and transfer the newly generated JPG files. 3) The large LCD screen gives a good enough image to ensure that the slide is centered before conversion. Although not documented in the manual, it is possible to change the size of the clipping aperture so that none of the image is lost if that is what you want. See the FAQs on Wolverine's web site. What does not work so well is the image conversion for bright areas. Other reviewers have noted that a yellow tint is produced for regions that should be bright white. I asked Wolverine technical support about this and was told that it is normal and that it can be reduced by lowering the EV value or corrected with post processing. I disagree and think it is either a flaw in the auto-scaling process or the JPEG compression process. You can see prominent yellow areas in PICT0280_19.jpg. This image was made with EV set to 0. Compare this to PICT0279_19.jpg which was made from the same slide with EV set to -2. The prominent yellow area is much whiter but still contains a subtle yellow tint. The dark areas are almost pure black and the image content is lost. Note that the image has been cropped identically in both files so as to eliminate the black border contained in the original scans. I used GIMP (GNU Image Manipulation Program) to analyze these two files to see what was going on and perhaps find a correction. The images PICT0280_19_RED.JPG, PICT0280_19_GRN.JPG, and PICT0280_19_BLU.JPG show histograms (log scale) for the red, green, and blue channels for PICT0280_19.jpg. The red and green channels appear typical, but the blue channel shows a picket fence structure above about the 65% of maximum value. This picket fence structure is not normal. Also, by inspecting the (red, green, blue) values for individual pixels, the yellow regions show values of (255, 255, 220) where red and green are at maximum (as expected for bright white), but blue is reduced thus giving a yellow cast. Moving slightly off of the yellow areas will give values like (252, 253, 253) where the blue channel shows higher values. It appears that whenever the red and green channels are at maximum, the blue channel is being reduced. So what happens when the EV is reduced before the conversion? The images PICT0279_19_RED.JPG, PICT0279_19_GRN.JPG, and PICT0279_19_BLU.JPG show histograms (log scale) for the red, green, and blue channels for PICT0279_19.jpg. The red and green channels again appear typical, but the blue channel shows a steep drop-off in pixel counts at intensities above about 85% of maximum value. This drop-off is not normal and accounts for the subtle yellow tint in the bright areas. For images taken with either EV set to 0 or with EV set to -2, the blue channel information is distorted at high intensities. Correcting this sort of error is difficult using the simple color correction capabilities available in most image manipulation software, including GIMP. The fundamental problem is that the image error is not distributed uniformly over the pixel intensities and there is no easy way to separate the pixels which need correction from those which don't. I sent a copy of this review to Wolverine and asked for their comments. I received a reply thanking for the review and with a statement that engineering will be pushed daily to find a proper solution and that firmware will be published ASAP to solve the issue. It is encouraging to see that Wolverine is acknowledging the problem and working to correct it. Still, I will hold off buying another one until I see evidence that the problem has been fixed.