The Client
Adaeze Obiechina is a commercial photographer based in Abuja. She specialises in corporate and advertising photography — the kind of work where a campaign shoot generates 2,000 frames and the client expects selects and retouched finals within 48 hours. Her clients include several major Nigerian banks, two telecommunications companies, and a growing roster of fashion and lifestyle brands that have discovered that professional Nigerian photographers can deliver at a standard that expensive expatriate photographers were previously brought in for.
She shoots on a Hasselblad X2D 100C — a 100 megapixel medium format digital camera. The files are enormous: a single RAW file is 200MB. A full shoot day generates 300–400GB of raw data. Processing this volume requires hardware that was genuinely designed for it. Adaeze's existing machine was not.
The Challenge
Her existing workstation was a desktop she'd bought in 2021 — a Core i7-10700K with 32GB DDR4 and an older NVIDIA GTX 1070. At the time, she was shooting on a Sony A7R IV (61MP, 60MB RAW files). The machine handled those files acceptably if not quickly. When she upgraded to the Hasselblad in 2025, the hardware became the bottleneck immediately.
Loading a single 200MB Hasselblad RAW file in Capture One took 38–44 seconds. Culling 400 images — scrolling through to select the best frames — required waiting 30–40 seconds between each image. A full-day shoot that should take 3–4 hours to cull took 9–10 hours. She had started arriving at shoots hoping for fewer frames because more frames meant more time behind a slow computer.
The GPU acceleration in Capture One was not helping because the GTX 1070 lacked the VRAM to cache multiple Hasselblad RAW previews simultaneously. The CPU was the second bottleneck — single-core decode performance determines how fast RAW files are decompressed, and the 10700K was not fast by 2026 standards.
The Consultation
Adaeze's workflow was Capture One Pro for culling, selection, and primary colour work, with occasional Photoshop for composite and retouching work. She printed large — up to A0 — so colour accuracy and consistent calibration mattered more than speed at the final output stage. She needed fast single-core CPU performance for RAW decode, substantial VRAM for caching previews, and the fastest possible NVMe storage for file access.
We also discussed her storage workflow. She was copying shoot files from CFexpress cards to an external USB-C SSD and then working from that drive — a single point of failure with no backup during the working window. We recommended a two-drive setup: primary fast NVMe for active work, mirrored backup drive for shoot originals, with the backup connected as soon as cards were ingested.
Her budget was ₦3.8 million. We built within it.
The Build
Photography Editing Workstation — ₦3.65 million:
- CPU: Intel Core i9-14900K — the fastest single-core performance available; RAW decode is heavily single-threaded
- RAM: 128GB DDR5-5600 — Capture One caches aggressively; more RAM means more previews held in memory
- GPU: NVIDIA RTX 4080 Super 16GB — 16GB VRAM for caching multiple 100MP previews simultaneously; Capture One GPU acceleration is well-optimised for NVIDIA
- Storage (primary): 4TB Samsung 990 Pro NVMe — file reads at 7,400 MB/s; Hasselblad RAW files load from this drive in under 2 seconds
- Storage (backup): 2× 8TB IronWolf in RAID-1 — shoot originals mirrored immediately on ingest
- Display: Calibrated 27" 4K IPS (BenQ SW271C with hardware calibration support) — Adaeze provided her own
- UPS: APC 2200VA — large enough to handle the RTX 4080 Super's draw during a rendering session
- Cooling: 360mm AIO — quiet, sufficient for the 14900K under sustained processing load
The Result
We ran benchmarks on delivery. Loading a single Hasselblad 100MP RAW file in Capture One from the NVMe: 1.8 seconds. Previously: 40 seconds. Culling 400 images — scrolling at a comfortable pace through the full shoot — now takes approximately 2 hours instead of 9. The GPU preview rendering fills in within a second of landing on each frame.
Adaeze described it as "a completely different job." She now culls the same day as the shoot rather than the following morning. For 48-hour client deliverable windows, that's the difference between comfortable and scrambling. She has taken on two additional shoot days per month since the build because the post-production no longer feels like a second job.
Photoshop retouching on 100MP files — layer operations that previously caused multi-second pauses — is effectively instant. She described opening a 12-layer composite file and having it respond "like a JPEG."
Key Takeaway
For high-resolution photography, the bottlenecks are specific and technical: single-core CPU speed for RAW decode, VRAM for preview caching, and NVMe speed for file access. A machine that excels in general benchmarks but falls short in these specific areas will still feel slow for this workflow. Getting these three things right transforms culling and editing from a grind into something that keeps pace with your creative process.
If you shoot medium format or high-resolution full-frame and your editing feels sluggish, the machine is the problem. And the machine is solvable.
Are you a photographer or visual creative in Nigeria? See the Creator Series or talk to our team about a workstation built for your specific camera and workflow.